JPH08330336A - Pickup device and chip positioning method - Google Patents

Abstract

PURPOSE: To accurately position all chips in the center of the visual field of camera by a method wherein the chips are arranged in the center of visual field of camera by giving the correction amount obtained by the preceding recognition to the preceding recognition data. CONSTITUTION: An image processing part 3 pre-recognizes a plurality of chips T within the visual field (m) of camera while the chips T are in operation, the image processing part 3 recognizes the amount of deviation of the center (c) when the chips T are arranged on the center (c) within the camera's visual field (m) based on the above-mentioned pre-recognition, an automatic teaching, in which the amount of deviation-recognized is used as the correction amount of chip position of pre-recognition, is conducted, and when the chip of the same position is pre-recognized while the correction amount by the automatic teaching is being renewed, the correction amount obtained by the last pre-recognition is given to pre-recognition. On the control part 7, a movable table 2 is operated by a movable table driving part 6 based on the output of the image processing part 3, and the chip T is a arranged on the center C. As a result, highly reliable high-speed pickup can be accomplished.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a pick-up device and a chip positioning method, and more particularly to a pick-up device used in the manufacture of electronic parts and a chip positioning method for recognizing and positioning the chip when picking up the chip. .

[0002]

2. Description of the Related Art For example, in a mounting step in the manufacture of a semiconductor device, a semiconductor chip (hereinafter referred to as a chip) divided from a semiconductor wafer and aligned is image-recognized, and a good chip is picked up.

Specifically, an optical system (including a camera) picks up images of chips arranged in an XY state on a movable table in an aligned state, and a plurality of chips within the picked-up camera field of view are subjected to image processing. In this image processing, the image data of the chip obtained from the video signal by the optical system is stored in the frame memory. A non-defective chip is selected by image recognition based on the image data and positioned at the pickup position based on the recognition data. Then, a non-defective chip positioned at the pickup position is vacuum-sucked by a collet to be picked up and mounted on an island of the lead frame.

Specifically, the image processing is executed in a state where the center of the camera field of view is set to coincide with the pickup position. For example, as shown in FIG. The image data of the chip T obtained by binarizing with the threshold value L is stored in the frame memory. Based on this image data, as shown in the same figure, the midpoint between the rising edge a and the falling edge b of the binarized signal A is set as the center position of one chip T (hereinafter referred to as the chip position), and the camera field of view is set. The chip position is image-recognized with respect to the center c. As a result, the distance d from the center c of the camera field of view to the chip position becomes the distance required for movement. In the figure, the chip T (shown by the broken line) located on the left side of the center c of the field of view of the camera may not actually exist because it has already been picked up, but it will be described as being virtually present. .

Based on the chip position thus obtained by image recognition, if the chip is a non-defective product, the chip at the chip position is set to the center c of the camera visual field by the movement of the movable table (distance d). Position and arrange.
After that, the non-defective chip positioned at the pickup position which is the center c of the camera field of view is picked up by the collet.

[0006]

By the way, as described above, the chips are picked up by the optical system, a plurality of chips in the field of view of the picked-up images are subjected to image processing, and the images are obtained by the image recognition based on the image data. With the tip position,
By moving the movable table, the chip at the chip position is positioned and arranged at the center of the visual field of the camera.

However, in the field of view of a camera that images a plurality of chips, brightness and distortion may occur due to the illuminance distribution on the chip surface, lens distortion of the optical system, etc., and overall uniformity may be impaired. For example, due to the above-mentioned illuminance distribution on the chip surface, lens distortion of the optical system, etc., the video signal S ′ by the optical system does not become a regular rectangular wave as shown in FIG. When the video signal S ′ is binarized by a predetermined threshold value L when the waveform distortion generated in the chips located at both ends is generated, the binarized signal A ′ is located in the chips located at both ends of the camera visual field. The rising edge a'due to the above shifts from the actual rising edge a (see FIG. 3), and the chip center, which is the midpoint between the rising edge a'and the falling edge b, shifts from the actual chip center.

Thus, when the chip center obtained by the binarized signal A'is deviated from the actual chip center,
Even if the chip T is moved (distance d ′) to be positioned at the center c of the camera visual field, the distance from the actual chip center to the center c of the camera visual field is d (d ′ <d). Since the position is displaced from the center c, it is necessary to accurately position the chip center again in the center c of the camera visual field, which increases the time for positioning the chip.

Therefore, the present invention has been proposed in view of the above problems, and an object of the present invention is not to be influenced by the illuminance distribution on the chip surface, the lens distortion of the optical system, or the like. It is an object of the present invention to provide a pickup device and a chip positioning method capable of accurately positioning the lens in the center of the camera visual field.

[0010]

As a technical means for achieving the above object, the method of the present invention is such that a plurality of chips are recognized in advance in the camera visual field in which the chips arranged in a plane in the XY direction are imaged in an aligned state. However, when the chip to be picked up is positioned and arranged in the center of the camera visual field based on the preceding recognition data, when the chip to be picked up is arranged in the center of the camera visual field based on the preceding recognition data. Re-recognize the amount of deviation from the center and use it as the correction amount for the chip position that was previously recognized,
At the time of prior recognition of a chip at the same chip position, the correction amount obtained by the previous prior recognition is added to the prior recognition data to arrange the chip at the center of the camera visual field.

Note that automatic teaching is performed by re-recognizing the amount of deviation from the center when the chip to be picked up is arranged at the center of the camera visual field based on the prior recognition data, and using it as the correction amount of the previously recognized chip position. Can be performed, and the correction amount by the automatic teaching can be updated for each prior recognition.

Further, in the device of the present invention, the chips arranged in a plane on the movable table in the XY directions in an aligned state are imaged by the optical system for image recognition, and the chips selected based on the recognition data are recognized by the arm system. In a device that sequentially picks up, a plurality of chips in the camera visual field imaged by the optical system are recognized in advance, and the amount of deviation from the center when the camera is arranged in the center of the camera visual field is recognized again based on the preceding recognition data. , Automatic teaching is performed with the correction amount of the tip position that was previously recognized, and the correction amount by the automatic teaching is updated for each prior recognition. An image processing unit that adds the obtained correction amount to the preceding recognition data, and the movable table is activated based on the output of the image processing unit. Characterized by comprising a control section for arranging the chip on the center of the camera field of view.

[0013]

According to the present invention, a plurality of chips are recognized in advance in the camera field of view, and the amount of deviation from the center when the chip to be picked up is placed in the center of the camera field of view is recognized again based on the preceding recognition data. It is used as the correction amount of the chip position that is recognized in advance, and when the chip at the same chip position is recognized in advance, the correction amount obtained in the preceding recognition is added to the recognition data and the chip is placed in the center of the camera field of view. To do.

By doing so, the chip center obtained from the image data is displaced from the actual chip center due to the waveform distortion of the video signal from the optical system due to the illuminance distribution on the chip surface, the lens distortion of the optical system, and the like. However, the positional deviation can be corrected by the correction amount, and the chip is accurately positioned at the pickup position which is the center of the camera visual field based on the preceding recognition data and the correction amount.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described with reference to FIGS.

First, the pickup device of the present invention has the following schematic structure. That is, as shown in FIG. 1, an optical system 1 including a camera takes an image of a chip T which is arranged on a movable table 2 in a plane in the XY directions, and a video signal from the optical system 1 is imaged by an image processing unit described later. In 3, image processing is performed by binarization or the like. Image data obtained by image processing of the chip T in the camera view m (see FIG. 2) captured by the optical system 1 is stored in a frame memory, and a good chip T is selected by image recognition based on the image data. Then, the movable table 2 is driven and positioned according to the recognition data at the pickup position which is the center c of the camera visual field m.
Then, the good chip T positioned at the pickup position is vacuum-sucked by the collet 5 of the arm system 4 and picked up.

The image processing unit 3 performs image processing by converting the video signal from the optical system 1 described above into a binarized signal or the like based on the camera view m of the plurality of chips T. The movable table drive unit 6 drives the movable table 2 based on an output signal from the control unit 7, which will be described later, to position the non-defective chip T at the pickup position, and the pickup drive unit 8 sets the pickup position to the pickup position. The one chip T located at is picked up by vacuum suction by the collet 5.

Specifically, the image processing section 3 recognizes a plurality of chips T in the camera field of view m in advance during the pickup operation of the chip T, and based on the preceding recognition data, the center of the camera field of view m. While re-recognizing the amount of deviation from the center c when it is placed at c, the automatic teaching is executed by using it as the correction amount of the previously recognized chip position, and the correction amount by the automatic teaching is updated for each prior recognition,
At the time of prior recognition of the chip T at the same chip position, the correction amount obtained in the previous prior recognition is added to the prior recognition data. In the control unit 7, the movable table 2 is operated by the movable table driving unit 6 based on the output of the image processing unit 3, and the chip T is arranged at the center c of the camera visual field m.

The chip positioning and the pick-up operation by the pick-up device having the above-mentioned structure are specifically as follows.

First, a plurality of chips T on the movable table 2
Is imaged by the optical system 1, and the video signal is captured by the image processing unit 3
Image processing is performed by binarization, and image recognition is performed based on the image data. At this time, the image processing unit 3 precedently recognizes a plurality of chips T in the camera visual field m. This prior recognition is performed during the pick-up operation of the chip T, and the chip T of the chip T is detected based on the presence / absence of the chip itself, the presence / absence of a defective appearance such as a chip, and the presence / absence of a defective mark based on the characteristic inspection result applied in the previous process. The quality of the image is determined by the image processing unit 3. Based on the output of the image processing unit 3, a non-defective chip T is selected from the determination result, and in the control unit 7, the movable table 2 is driven by the movable table driving unit 6 and the non-defective chip T described above is recognized by the camera based on the recognition data. Center c of field of view m
Position it at the pickup position.

At this time, the chip center at the pickup position and the center c in the camera visual field m may deviate from each other. That is, as shown in FIG. 4, the image signal S ′ by the optical system 1 does not become a regular rectangular wave due to illuminance distribution on the chip surface, lens distortion of the optical system, etc. Waveform distortion) that occurs in the chip located at the position a, the rising edge a'of the chips T located at both ends of the camera field m of the binarized signal A'which is obtained by binarizing the video signal S'with the threshold value L. The cause is that the midpoint between the rising edge a ′ and the falling edge deviates from the actual chip edge (see FIG. 3) and deviates from the actual chip center.

Therefore, in the present invention, when the center of the chip obtained by the binarized signal A'is deviated from the actual center of the chip as described above, the chip T is set based on the instruction from the control unit 7. The image processing unit 3 re-recognizes the amount of deviation between the center of the chip and the center c of the camera visual field m, which occurs when the positioning is performed at the center c of the camera visual field m. The automatic teaching is executed with the amount of deviation between the center of the chip obtained by this re-recognition and the center c of the camera field of view m as the correction amount of the previously recognized chip position. As a result, for each chip T, the correction amount based on the shift amount is added to the preceding recognition data.

By updating the correction amount by the automatic teaching for each preceding recognition, the influence due to a change with time is suppressed as much as possible, and the latest correction amount is always added to the preceding recognition data. At the time of prior recognition of the chip T at the chip position, the correction amount obtained in the previous prior recognition is added to the prior recognition data, and the movable table drive unit 6 drives the movable table 2 based on a command from the control unit 7. Then, based on the recognition data, the chip T is accurately positioned and arranged at the pickup position which is the center c of the visual field m of the camera.

The correction amount by the above-mentioned automatic teaching is the center of the chip and the center c of the camera field of view m generated when the chip T is positioned and arranged at the center c of the camera field of view m.
The amount of deviation from the above is used as a correction amount when the position of the chip T in the camera field of view m is recognized as it is in pixel units, or is held as the screen resolution when the position of the chip T in the camera field of view m is recognized. do it.

[0025]

According to the present invention, a plurality of chips in the field of view of a camera are recognized in advance by an image processing unit, and the amount of deviation from the center when the camera is arranged in the center of the field of view of the camera is re-recognized based on the preceding recognition data. Recognize and execute the automatic teaching with the correction amount of the tip position that was recognized in advance, and update the correction amount by the automatic teaching for each prior recognition. Since the chip is arranged in the center of the camera field of view by adding the correction amount obtained by the recognition to the preceding recognition data, without being adversely affected by the illuminance distribution on the chip surface or the lens distortion of the optical system, Since each chip can always be accurately positioned at the pickup position, which is the center of the camera's field of view, a reliable and high-speed pickup can be achieved. Its practical value made the current is large.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of a pickup device of the present invention.

FIG. 2 is an explanatory view showing a camera field of view in prior recognition in the embodiment of the present invention.

FIG. 3 is a waveform diagram showing a normal video signal and a binarized signal obtained by binarizing the video signal with a threshold value.

FIG. 4 is a waveform diagram showing a distorted video signal and a binarized signal obtained by binarizing the video signal with a threshold value.

[Explanation of symbols]

 T chip m camera field of view c camera field of view center 1 optical system 2 movable table 3 image processing unit 4 arm system 7 control unit

Claims (4)

[Claims] 1. A camera field of view in which a plurality of chips are recognized in advance in a camera field of view in which chips arranged in a plane in the XY directions in an aligned state are sensed, and a chip to be picked up is sensed based on the pre-recognition data. At the time of positioning and arranging at the center of the center, the amount of deviation from the center when the chip to be picked up is arranged at the center of the camera field of view based on the preceding recognition data is re-recognized, and this is taken as the correction amount of the previously recognized chip position. A chip positioning method characterized in that, at the time of prior recognition of a chip at the same chip position, the correction amount obtained in the previous prior recognition is added to the prior recognition data to arrange the chip at the center of the camera visual field. 2. Automatic teaching in which the amount of deviation from the center when the chip to be picked up is placed at the center of the camera field of view based on the prior recognition data is recognized again as the correction amount of the chip position that was previously recognized. The chip positioning method according to claim 1, further comprising: 3. The chip positioning method according to claim 2, wherein the correction amount by the automatic teaching is updated for each prior recognition. 4. An apparatus for picking up an image of a chip, which is arranged in a plane in the XY direction in an aligned state on a movable table by an optical system, for image recognition, and sequentially picking up a chip selected based on the recognition data by an arm system. , A plurality of chips in the camera field of view picked up by the optical system are recognized in advance, and the amount of deviation from the center when the camera is arranged in the center of the camera field of view is recognized again based on the preceding recognition data, and it is recognized in advance. Automatic teaching is performed as the correction amount of the specified chip position, and the correction amount by the automatic teaching is updated for each prior recognition. Is added to the preceding recognition data, and the movable table is operated based on the output of the image processing unit to operate the chip in the field of view of the camera. And a control unit disposed at the center of the pickup device.